The first spatio-temporal analysis of foraging flights of wild rock doves is here reported. Birds living in one of the last relatively pure demes of wild Columba livia in the Mediterranean region (Capo Caccia, NW Sardinia) were observed on their foraging routes and in the course of daily movements to and from the colonial site. Foraging sites were located by systematic searches of colour-marked birds in the study area.The pigeons followed rather stable routes in reaching foraging sites, which are mainly located in agricultural fields up to around 19 km from the colonial site. In autumn, doves tended to reduce the extent of foraging flights, feeding almost exclusively on lentiscus fruits located near the colony. Foraging flights in spring and summer differed from those in autumn and winter in the patterns of daily activity, the number of birds moving, and the sizes of the flocks. The possible relationships between the spatio-temporal characteristics of foraging flights and other aspects of the biology of rock doves, especially in the breeding season, are discussed.KEY WORDS: Rock dove -Columba l. livia -Foraging flightsActivity pattern. ACKNOWLEDGEMENTS

Previous experiments on wild rock doves ( Columba l. livia Gmelin) released within their familiar area revealed an evident effect of fast-shifting, although shifted doves, but not controls, tended to orient homeward. Such an outcome suggested a possible influence of the release time per se on the directional choices of the tested doves. In the present study, this hypothesis was investigated by comparing the orientation of slow-shifted birds to that of two control groups released at different times of the day. As would be expected if doves make use of a time-compensated sun compass, the bearings of shifted birds were deflected clockwise with respect to controls. The time of release itself seemed to influence only the scatter of the vanishing bearings of controls. These findings testify to the substantial similarity of clock-shift effects on the initial orientation of rock doves to those on homing pigeons released from familiar sites.

The homing behaviour of free-living birds has rarely been systematically investigated, and so it is difficult to give a satisfactory comparative picture. Data are here presented on the initial orientation of rock doves displaced inside and outside the area that they daily cross during their foraging flights. Inside the familiar area, the results show marked variability in the directional choices of the released birds, probably related to temporal variability in the location of food patches during the year. Indeed, in autumn, when the doves feed almost exclusively near the colony site, they tend to fly home, but in summer, when food patches are dispersed over a wide area around the colony, this behaviour is not evident. Outside the familiar area, no clear evidence of a homeward directedness is obtained. Two releases inside the familiar area with clock-shifted birds produced ambiguous results and it is difficult to attribute the behaviour of the animals to a change in their motivation to reach home in different times of the day, or to the use of a time-compensated sun compass. On the whole, the presented results point to the difficulties of comparing rock dove behaviour to that of the homing pigeon, and show the importance of considering the influence of ecological and motivational variables in evaluating the orienting performances of wild birds.

Data on the effect of repeated releases on the homing behaviour of sand martins released four times from the same site (beyond 50 km from the nesting colony) are presented. The results, obtained by two series of releases, show a progressive improvement of both the initial orientation and the homing performances. When the birds released four times from the same site were displaced in a new site, located roughly at the same distance, but in the opposite direction to the former, they do not change directional preferences with respect to the last experimental release; the homing speeds were also not different to those recorded for birds without any release experience. While the learning of orienting cues can not be excluded, the data point out that the habituation to the handling stress and a form of directional training seem to play a major role in determining the observed behaviour. The familiarity with a site does not seem to improve the homing behaviour of sand martins when released from a different one.

No abstract